Looking for treatment and prevention requires a good way to measure if a therapy is working. Using clinical progression to Alzheimer's disease (AD) requires a huge multi-year study to get any real statistical power. Not everyone goes on to develop AD, people die from other stuff, etc. If a treatment doesn't work, you've just wasted lots of $ and time to find that out (e.g., http://www.ncbi.nlm.nih.gov/pubmed/18305231). Maybe you had your dose wrong, maybe you had the timing off,... The search space for a treatment is HUGE, there has a to be an efficient way to quickly (relative here) and accurately determine if a therapy works. Having a way to detect and monitor neurodegenerative diseases would be awesome from a research standpoint. It would allow therapy to be tested using a cross sectional study rather than a longitudinal study.
Wife and I recently signed up for Simple Mobile. It works with TMobile or unlocked GSM phones and is $40 for unlimited talk/text/data (ok, data is probably not really unlimited, but enough for my basic needs). I've only had about 10 days but seems fine. I also found that I can buy the plan online (I used a place called pinzoo) and then avoid tax. May not be best for you since you really only want texting.
Where do you publish that you do not check galley proofs? I too have had a few articles published and am always forced to approve the galley proofs before they document goes to press. Maybe there are journals that don't require this, but I know many have it mandatory step. If you choose to blindly accept the proof without changes that reflects more on you than the publisher. Not submitting revisions either means you write perfect and no reviewer/editor has any questions/comments (congratulations if this is the case) or that you are able somehow avoid the revisions.
I do agree with you that its tough to see the value added by journals from the journals other than the perceived clout they carry. From a CV standpoint, its "better" to publish in IEEE TMI than slimak's wonderful world of science. But, that is only because of the weight our peers assign the journals and really not a real value (to me).
I agree that relying on spell check is bad, but I don't know a single person that turns off spell/grammar checks. I'm sure they exist, but I don't see anything wrong with spell check. Catching all typos in proofreading is difficult for any moderately long text (I'll probably have at least one in this ultra-short text). Having done technical and scientific writing for many years I still prefer LaTeX over Word with two exceptions: Word does nice enough change tracking for collaborating/revising and spell check saves a bit of time.
I've seen this a few times lately and am curious why this belief is held. Maybe (probably) I'm missing something but I would think that source code would be an asset and potentially valuable in a few cases:
1) A complex system that took significant time to develop. Something like MS Word. While it may not be your favorite it certainly is an assest and has a value. A word processor is easy to think of, but Word is difficult/time consuming to implement (I'm guessing).
2) Software the implements a trade secret. Something like an auto stock trading system or the Google search results ranking algorithm. Again, you may hate these and they are of no value to you, but if your livelihood was on the line would you want to release the source?
I completely agree that the source code to a generic sorting algorithm of your favorite memory copy routine has no value, but even and AC must see there are exceptions. Of course, I could just be stupid.
In the US virtually all x-ray machines (including medical) are operated by un-certified radiologists. Radiologists interpret the images, they do not (typically) run the imaging devices. Radiographer or radiologic technologist (or just "tech" as they are typically called in the field) run the devices. Fortunately, the techs in medicine are typically well trained and certified. I'm not sure about the TSA team, but probably not so much. So your overall point is probably still accurate.
From what I recall the GRE also does (or did 10 years ago) a similar adaption. Sounds like the CPA exam is similar. As a side benefit of such adaption you can somewhat tell how you are doing. If the test is easy you are doing either very well because you are so super smart, or very poor because you fall in the less-desirable part of the intelligence bell curve!
Wouldn't it be better to wait and see if they do fail (which can be detected with 100% accuracy without EEG) and adapt the question then? Who could stay engaged when questions are changing while they are working on them?
That is BS in general. There are certainly some teachers that this applies to but any parent can request an observation to see exactly what is being done in the classroom. If you to examine you can. A teachers job is to teach the kids, not show the parents what is being taught. If you want to know what they are doing, go and check it out or ask the teacher outright. I am not a teacher, but have always found the district my child attends to be open and helpful.
I agree that writing and typing a good skills today and I really had using touchscreens and the tiny keyboards on mobile devices, but we have to adapt. We used to record history on stone or clay using hieroglyphs, I'm really glad that fad passed. I sincerely hope that a few hundred years from know our descendants will think of us as Neanderthals that had to didn't even had neural implants (or something even more amazing that I am too primitive to even dream of).
Most scientific journals that I have experience with to not pay authors in any way. This is certainly the case with all IEEE journals and several other scientific journals. Signing over the copyright is the cost of entry if you want your work published. There are probably exceptions to this, possibly for work that is easily identified as ground breaking. But my experience has always been that there is nothing paid when the copyright is transfered. In fact, most journals still ask for printing charges. This are usually optional (and I opt out) except when color figures are included in the manuscript. If there are journals paying authors I would like to know.
Edison has gotten far more coverage in the history books (at least US ones), He was probably best at business, although he is known as an inventor. On the other hand, Tesla was, without a doubt, the greatest engineer that has ever lived. He is proof that a formal advanced education is not necessary for scientific greatness. It is too bad that most people don't realize the impact he truly had.
If you finish grad school with a GPA of less than 3.5 there is a good chance you were not studying something you were passionate about (particularly for a doctoral degree). The bell curve does not exist in (good) grad programs when it comes to grades. This is because a disproportionate number of students are outstanding making for a lot of top marks. The grad engineering courses I took (possibly excluding "trial by fire courses") typically had an B+/A- average and grades amounted to: A=solid grasp of concepts, B=partial understanding, consider re-taking or serious reading, C=fail (did you even buy the text?). I never new of a single D given out, F was only in extreme cases (student though class was dropped).
Naturally, this may vary by program, department, and university. But I know that it is relatively common in engineering.
That said, even the academically "poor" students were often very gifted when it came to research. Ultimately, publications and products make for placement if you want an academic/research life after grad school. A respectable GPA may be be necessary for corporate jobs, but will probably carry less weight for faculty/post-doc type job searches. Nothing helps you make the short list better than proven research skills demonstrated a few first author publications.
A high resolution signal cannot be created from a low resolution signal. Period*. The best you can do is get more sample points of the original low resolution signal. That said, upconverted signals do look better (to me at least) than standard def, but there is no additional information.
* If you impose additional constraints on the data during the interpolation, then you may be able to do slightly better. This falls in the realm of sub-Nyquist sampling.
I have a PhD, but my wife always says jokingly I am not a "Real Doctor" because I cannot treat patients or prescribe meds. I don't know how it is outside the US, but almost everyone outside of academics that I encounter thinks MD when they hear someone is a Doctor.
Although watermarking is not my field of expertise, I often attend the watermarking sessions at research conferences because I find them interesting. Current state-of-the art techniques are designed to embed watermarks such that they can survive a significant number of operations including image manipulations and re-compression. It comes down to the fact that the watermark stored is MUCH smaller than the amount of data that can be added to the image (without introducing visually objectionable artifacts). Because of this, the watermark data is redundantly stored and includes error-correction coding (I think the watermarking people call it something else though).
A good watermark will be very hard to remove/destroy without making the source image worthless. From what I have seen, watermarks can often survive image retouching and format conversion but are venerable to image distortions (sheers, warps, stretch, etc).
This is pretty standard stuff in almost any junior/senior level signal or image processing course. Whether the operation can be undone depends heavily on the type of blurring. For example, if the blurring operation is linear then it can be represented by a matrix operation. This means we can write the blurred image (y) in terms of the blurring kernel matrix (H) and the original image (x) as: y=Hx (assume x and y are column vectors of all image samples). This operation can be inverted perfectly as long as H is a non-singular matrix as: x_hat = inv(H)*y = inv(H)*H*x = x.
When H is singular or non-rectangular the typical approach would be to make x_hat = inv(H^T*H)*H^T*y, which is the least-squares approximation of x from the blurred observation y (^T denotes matrix transpose).
When the operation is non-linear it generally get MUCH harder. Here, the most common approach would be to use some statistically optimal method or incorporate a priori image information.
Why is a great question. I will be upgrading to an Intel-based mac next year and It will be a cold day in hell before Windows gets anywhere near it. I switched to Macs a few years back to get away from Windows.
I guess it could be nice for users that need certain Windows apps -- however I'm afraid that current applications offered for both OS X and Windows will slowly drop OS X support to save on devleopment costs. That would suck.
Linear Algebra is a fundamental topic like calculus, classic physics, or discrete math. Depending on how it is taught, it can be very easy or fairly challenging. You will spend a good deal of time in the beginning on vector spaces -- these seem pretty useless until you get farther into the class and see their power. A solid understanding will help you apply the concepts of basis change and projection onto a subspace to real problems. Overall, linear algebra is a very rich and useful topic. enjoy.
As an apple customer, I can tell you that I don't give a crap what kind of chips they use as long as 1) I can get a faster powerbook 2) The cost does not go (any) higher 3) Any migration is relatively painless 4) The products still look cool
actually, now that I think of it, I really only care about #4
Slashdot Mirror
Looking for treatment and prevention requires a good way to measure if a therapy is working. Using clinical progression to Alzheimer's disease (AD) requires a huge multi-year study to get any real statistical power. Not everyone goes on to develop AD, people die from other stuff, etc. If a treatment doesn't work, you've just wasted lots of $ and time to find that out (e.g., http://www.ncbi.nlm.nih.gov/pubmed/18305231). Maybe you had your dose wrong, maybe you had the timing off, ... The search space for a treatment is HUGE, there has a to be an efficient way to quickly (relative here) and accurately determine if a therapy works. Having a way to detect and monitor neurodegenerative diseases would be awesome from a research standpoint. It would allow therapy to be tested using a cross sectional study rather than a longitudinal study.
Wife and I recently signed up for Simple Mobile. It works with TMobile or unlocked GSM phones and is $40 for unlimited talk/text/data (ok, data is probably not really unlimited, but enough for my basic needs). I've only had about 10 days but seems fine. I also found that I can buy the plan online (I used a place called pinzoo) and then avoid tax. May not be best for you since you really only want texting.
Where do you publish that you do not check galley proofs? I too have had a few articles published and am always forced to approve the galley proofs before they document goes to press. Maybe there are journals that don't require this, but I know many have it mandatory step. If you choose to blindly accept the proof without changes that reflects more on you than the publisher. Not submitting revisions either means you write perfect and no reviewer/editor has any questions/comments (congratulations if this is the case) or that you are able somehow avoid the revisions.
I do agree with you that its tough to see the value added by journals from the journals other than the perceived clout they carry. From a CV standpoint, its "better" to publish in IEEE TMI than slimak's wonderful world of science. But, that is only because of the weight our peers assign the journals and really not a real value (to me).
I agree that relying on spell check is bad, but I don't know a single person that turns off spell/grammar checks. I'm sure they exist, but I don't see anything wrong with spell check. Catching all typos in proofreading is difficult for any moderately long text (I'll probably have at least one in this ultra-short text). Having done technical and scientific writing for many years I still prefer LaTeX over Word with two exceptions: Word does nice enough change tracking for collaborating/revising and spell check saves a bit of time.
I've seen this a few times lately and am curious why this belief is held. Maybe (probably) I'm missing something but I would think that source code would be an asset and potentially valuable in a few cases:
1) A complex system that took significant time to develop. Something like MS Word. While it may not be your favorite it certainly is an assest and has a value. A word processor is easy to think of, but Word is difficult/time consuming to implement (I'm guessing).
2) Software the implements a trade secret. Something like an auto stock trading system or the Google search results ranking algorithm. Again, you may hate these and they are of no value to you, but if your livelihood was on the line would you want to release the source?
I completely agree that the source code to a generic sorting algorithm of your favorite memory copy routine has no value, but even and AC must see there are exceptions. Of course, I could just be stupid.
In the US virtually all x-ray machines (including medical) are operated by un-certified radiologists. Radiologists interpret the images, they do not (typically) run the imaging devices. Radiographer or radiologic technologist (or just "tech" as they are typically called in the field) run the devices. Fortunately, the techs in medicine are typically well trained and certified. I'm not sure about the TSA team, but probably not so much. So your overall point is probably still accurate.
From what I recall the GRE also does (or did 10 years ago) a similar adaption. Sounds like the CPA exam is similar. As a side benefit of such adaption you can somewhat tell how you are doing. If the test is easy you are doing either very well because you are so super smart, or very poor because you fall in the less-desirable part of the intelligence bell curve!
Wouldn't it be better to wait and see if they do fail (which can be detected with 100% accuracy without EEG) and adapt the question then? Who could stay engaged when questions are changing while they are working on them?
That is BS in general. There are certainly some teachers that this applies to but any parent can request an observation to see exactly what is being done in the classroom. If you to examine you can. A teachers job is to teach the kids, not show the parents what is being taught. If you want to know what they are doing, go and check it out or ask the teacher outright. I am not a teacher, but have always found the district my child attends to be open and helpful.
I agree that writing and typing a good skills today and I really had using touchscreens and the tiny keyboards on mobile devices, but we have to adapt. We used to record history on stone or clay using hieroglyphs, I'm really glad that fad passed. I sincerely hope that a few hundred years from know our descendants will think of us as Neanderthals that had to didn't even had neural implants (or something even more amazing that I am too primitive to even dream of).
Most scientific journals that I have experience with to not pay authors in any way. This is certainly the case with all IEEE journals and several other scientific journals. Signing over the copyright is the cost of entry if you want your work published. There are probably exceptions to this, possibly for work that is easily identified as ground breaking. But my experience has always been that there is nothing paid when the copyright is transfered. In fact, most journals still ask for printing charges. This are usually optional (and I opt out) except when color figures are included in the manuscript. If there are journals paying authors I would like to know.
Edison has gotten far more coverage in the history books (at least US ones), He was probably best at business, although he is known as an inventor. On the other hand, Tesla was, without a doubt, the greatest engineer that has ever lived. He is proof that a formal advanced education is not necessary for scientific greatness. It is too bad that most people don't realize the impact he truly had.
If you finish grad school with a GPA of less than 3.5 there is a good chance you were not studying something you were passionate about (particularly for a doctoral degree). The bell curve does not exist in (good) grad programs when it comes to grades. This is because a disproportionate number of students are outstanding making for a lot of top marks. The grad engineering courses I took (possibly excluding "trial by fire courses") typically had an B+/A- average and grades amounted to: A=solid grasp of concepts, B=partial understanding, consider re-taking or serious reading, C=fail (did you even buy the text?). I never new of a single D given out, F was only in extreme cases (student though class was dropped).
Naturally, this may vary by program, department, and university. But I know that it is relatively common in engineering.
That said, even the academically "poor" students were often very gifted when it came to research. Ultimately, publications and products make for placement if you want an academic/research life after grad school. A respectable GPA may be be necessary for corporate jobs, but will probably carry less weight for faculty/post-doc type job searches. Nothing helps you make the short list better than proven research skills demonstrated a few first author publications.
A high resolution signal cannot be created from a low resolution signal. Period*. The best you can do is get more sample points of the original low resolution signal. That said, upconverted signals do look better (to me at least) than standard def, but there is no additional information. * If you impose additional constraints on the data during the interpolation, then you may be able to do slightly better. This falls in the realm of sub-Nyquist sampling.
I have a PhD, but my wife always says jokingly I am not a "Real Doctor" because I cannot treat patients or prescribe meds. I don't know how it is outside the US, but almost everyone outside of academics that I encounter thinks MD when they hear someone is a Doctor.
http://www.imdb.com/title/tt0364343/
Although watermarking is not my field of expertise, I often attend the watermarking sessions at research conferences because I find them interesting. Current state-of-the art techniques are designed to embed watermarks such that they can survive a significant number of operations including image manipulations and re-compression. It comes down to the fact that the watermark stored is MUCH smaller than the amount of data that can be added to the image (without introducing visually objectionable artifacts). Because of this, the watermark data is redundantly stored and includes error-correction coding (I think the watermarking people call it something else though). A good watermark will be very hard to remove/destroy without making the source image worthless. From what I have seen, watermarks can often survive image retouching and format conversion but are venerable to image distortions (sheers, warps, stretch, etc).
This is pretty standard stuff in almost any junior/senior level signal or image processing course. Whether the operation can be undone depends heavily on the type of blurring. For example, if the blurring operation is linear then it can be represented by a matrix operation. This means we can write the blurred image (y) in terms of the blurring kernel matrix (H) and the original image (x) as: y=Hx (assume x and y are column vectors of all image samples). This operation can be inverted perfectly as long as H is a non-singular matrix as: x_hat = inv(H)*y = inv(H)*H*x = x.
When H is singular or non-rectangular the typical approach would be to make x_hat = inv(H^T*H)*H^T*y, which is the least-squares approximation of x from the blurred observation y (^T denotes matrix transpose).
When the operation is non-linear it generally get MUCH harder. Here, the most common approach would be to use some statistically optimal method or incorporate a priori image information.
Why is a great question. I will be upgrading to an Intel-based mac next year and It will be a cold day in hell before Windows gets anywhere near it. I switched to Macs a few years back to get away from Windows.
I guess it could be nice for users that need certain Windows apps -- however I'm afraid that current applications offered for both OS X and Windows will slowly drop OS X support to save on devleopment costs. That would suck.
Or, the author is from a country that uses the comma as a decima separator: http://en.wikipedia.org/wiki/Decimal_separator
Linear Algebra is a fundamental topic like calculus, classic physics, or discrete math. Depending on how it is taught, it can be very easy or fairly challenging. You will spend a good deal of time in the beginning on vector spaces -- these seem pretty useless until you get farther into the class and see their power. A solid understanding will help you apply the concepts of basis change and projection onto a subspace to real problems. Overall, linear algebra is a very rich and useful topic. enjoy.
As shipped, my Nano can only be charged via USB.
But tax has already been paid on that money (assuming it was originally reported as income).
Shouldn't grade inflation help you get a higher GPA? Would you really be better off with a 2.3 GPA under a non-inflated system?
As an apple customer, I can tell you that I don't give a crap what kind of chips they use as long as
1) I can get a faster powerbook
2) The cost does not go (any) higher
3) Any migration is relatively painless
4) The products still look cool
actually, now that I think of it, I really only care about #4